Term
| When rendering using the hardware transform and lighting pipeline or vertex shaders,what does the GPU do? |
|
Definition
| it caches the transformed and lit vertices |
|
|
Term
| What does storing these vertices accomplish? |
|
Definition
| avoids recomputing the same values whenever a vertex is shared between multiple triangles, thus saving time |
|
|
Term
| How much does the post-TnL cache increase rendering performance? |
|
Definition
|
|
Term
| How do you take full advantage of the TnL? |
|
Definition
| You must re-arrange the triangle rendering order. |
|
|
Term
| What is the easiest way to re-arrange the triangle rendering order? |
|
Definition
| Use the NVTriStrip library |
|
|
Term
| How does the TnL buffer work? |
|
Definition
| FIFO - First in First Out |
|
|
Term
| How many vertices can the TnL store? |
|
Definition
| varies in size from 10 (actual 16) to 18(actual 24) |
|
|
Term
| Can non-indexed draw calls take advantage of the cache? |
|
Definition
|
|
